HD1960MM1APU
Preface by D Landau 1999
I have recreated this document
almost 40 years after it was written.
It’s a well written and worth reading for anyone interested in the early
Minuteman Missile. The Minuteman was the
first solid propellant intercontinental missile and constituted a huge jump
forward, relative to the prior liquid rocket engine systems, which could take
all day to launch – if then. Technology
was permitting advances and the Air Force wanted something that could
immediately launch (thus the name Minuteman) be stored underground (thus
development of silo launch) and accurately hit a target on the other side of
the Earth – with very special emphasis on reliability. Such a vehicle and capability had never
existed before. Autonetics was given
the contract primarily because of it’s Inertial Guidance system developed for
the Navaho missile, and then being used by the Navy to guide their missile
carrying submarines.
Lou and I were a part of Flight
Control which became part of Autonetics when that division was split away from
Space Division. Prior to Minuteman Lou
had been our “Hydraulics Pump” man, I operated our extreme temperature
hydraulic test facility where we were developing advanced hydraulic systems. During the Minuteman proposal Lou worked
with systems engineers Art Greer and Bill Stobel. We were all young and close
friends, thus I had many one-on-one conversations with them when they tried out
ideas or sought information about our lab experience. We ran tests on experimental hardware they were obtaining –
including evaluation of Flywheel System to power the pumps. (An idea I
suggested to Art at lunch one day – lab tests proved this would not work with
the technology at hand.) We also
discussed seals at some length, I told them face seals were much more
vulnerable to mechanic error than were shaft seals. Thus we had no face seals.
We all worked for Paris Stafford at
the time, who in turn worked for George Keller. Though we were all very young we were working on leading edge
stuff and had considerable first hand experience prior to Minuteman. George sent Paris and I off to night classes
at UCLA to keep up on Bearings and Lubricants – we were developing the Mach 3
Navaho RamJet ICBM at the time. The
idea of a multistage solid propellant missile didn’t come as a surprise to me
as I had attended presentations sponsored by the Society of Mechanical
Engineers in downtown LA – people from STL, later TRW presented studies and
found three stages were optimum.
We
had considerable experience on how to electronically control hydraulic
servo-actuators. Gary Collins,
“invented” the linear position transducer and left the company to produce it,
it was soon on all US aircraft and missiles.
We were working with MOOG to develop a “dry” servo torque motor to
greatly reduce vulnerability to contamination.
Our Flight Control organization was the most advanced in the new field
of electronic control of hydraulic systems.
George Keller, as head of the committee on hydraulics for avionics was
constantly bringing people from other companies through our lab. We had already shifted from vacuum tubes to
semiconductors for servo control. We
were doing testing for the B-70, including Jim Anderson’s advanced servo-actuator
design.
Upon
contract go ahead Lou and Art wrote the procurement specs for the pump and
actuator subassemblies. They were party to selecting subcontractors Ling Tempco
Vaught for servoactuators and Vickers for the APU’s. The program was expanding rapidly; Bill went into controls system
analysis, Art and Lou into the Project Office and I was moved from the Test Lab
to lead engineer for the Minuteman Hydraulic Controls. Art and Lou prepared and presented a paper on
the APU (Auxiliary Power Unit). And Lou
followed with this paper on the Nozzle Control Units.
Lou
as project engineer and I as lead engineer were in daily contact and played key
rolls in frequent meetings with subcontractors. Thus I was intimately involved with ever detail in the following
paper. I will add my comments at the
end and not interfere with the flow of Lou’s paper
Time
was to prove this system was extremely reliable – all that attention to detail really
paid off.
We
lost track of Art Greer, about 10 years later he left the company. Lou went on to become Assistant Program
manager with an office along Mahogany Row.
He’s still an avid bridge player and we visit at each Flight Control
Christmas party. Bill Strobel died of a
heart attach at a very young age having recently married and becoming a father
for the first time. Jim Anderson, Frank
Lettang and Eliott Buxton are among the few Flight Control fellows left from
those very early days of Minuteman I.
Frank did the thermal protection for the NCU’s and Buck succeeded in
seeing that MM Flight Control used a digital computer – another first.
Darrell Landau 1999
I. Introduction
The
MINUTEMAN missile is a three-stage, solid-propellant, second-generation
ICBM. MINUTEMAN will be stored
underground in scattered silos, ready at all times to be launched with a push
of a button. MINUTEMAN's much
publicized performance and destructive capability will undoubtedly make it one
of the most potent weapons in our defense arsenal. The USAF now expects MINUTEMAN to be operational approximately one
year ahead of the original target date.
The overall weapon system concept of
MINUTEMAN is one which makes reliability one of the most critical performance
considerations challenging the management, engineering, and manufacturing
skills of the contractors. MINUTEMAN,
stored in its silo, is inaccessible for maintenance or repair. It requires a capability for intermittent
automatic, go/no-go, prelaunch checkout.
These missile requirements are to be served by a simple field logistics
concept. It is time, however, which
imposes the greatest challenge to operational reliability, for MINUTEMAN is
designed to be stored in its silo, in instant readiness, for periods of up to
five years.
Only a high degree of reliability in
systems and components will meet the stringent disciplines of this design
philosophy. And, if these reliability
goals are not reached, MINUTEMAN's usefulness as a weapon system will be
greatly affected. In brief, it can be said
that MINUTEMAN is not a larger or more accurate missile; it is only better
insofar as its design philosophy and its reliability will carry it. If these goals are not reached, MINUTEMAN's
overall effectiveness as a weapon system is diminished.
This paper discusses the design of a
hydraulic system that is expected to provide the high-performance operational
characteristics with the high degree of reliability required by the MINUTEMAN
missile.
A. GENERAL
In flight, MINUTEMAN's Flight
Control System converts steering information into the required thrust vectoring
action. Steering information is
supplied by the missile's inertial navigator and its digital computer. The Flight Control System uses a Nozzle
Control Unit in each of the three stages to position the gimballed nozzles
controlling the prime thrust.
In the assembled missile, the Flight
Control System is inaccessible for maintenance, repairs, or even minor
adjustment. Any malfunction or
out-of-spec drift requires a major missile disassembly.
B. NOZZLE CONTROL UNIT
A Nozzle Control Unit (NCU) in each
of the three missile stages provides roll, yaw, and pitch control. Each NCU contains four electro-hydraulic
position servos and their associated peripheral equipment. The hydraulic system (Figs. 1 and 2) is
comprised of four servo actuators and an auxiliary hydraulic supply married to
a common platform which serves as:
1. Hydraulic manifold. The hydraulic subassemblies (servoactuators
and APS) are of the plug-in type and fluid is transmitted through porting
within the platform.
2. Component
and subassembly mounting structure.
3. Base structure for a fully-integrated,
closed-hydraulic system.
The
system is completely self-contained with no external supply equipment
required. The system is cleaned, filled,
and bled in a controlled manufacturing area.
Service and maintenance, as required, are accomplished under carefully
specified conditions in controlled depots.
Field maintenance is not permitted for these reasons:
1. Recent
studies conducted on several major USAF ballistic missile programs found that
approximately 50 percent of critical equipment failures resulting in mission
abort were caused directly by human error. [1]
2. System
cleanliness can best be assured when filling, bleeding, and servicing
operations are carried out only under controlled conditions. Since contaminants increase the force of
mortality of high-performance, close-tolerance hydraulic devices, this approach
was considered mandatory.
C. HYDRAULIC FLUID
A modified version of MIL,-H-5606
hydraulic fluid is presently under test and is expected to be used in the
operational system. This fluid is
basically a pure cut of saturated hydrocarbons as close to
Fig 1
Typical Nozzle Control Unit Hydraulic System Schematic
Fig 2
Nozzle Control Unit Assembly
being
physically and chemically inert as can be found. Where, then, do the problems associated with storage stability
stem from? The principal problem
results from impurities which are naturally present in the petroleum crude
and/or those which are deliberately added.
The most effective way of minimizing the naturally occurring compounds
of sulfur, nitrogen, and oxygen present in the fluid is through the careful
selection of the source of the petroleum crude. The aromatic content of the fluid influences the swell of rubber
compounds and, therefore, must be controlled.
Since the aromaticity can be effectively controlled through the
selection of the petroleum crude, we find that two major variables, not always
compatible, influence the selection of the base stock. These variables (not controlled in the
MIL,-H-5606 specification) are to be brought under control in the modified
version of the specification.
Compounds deliberately added to
cause desirable properties are used only where necessary. In this respect, the polymethacrylate
thickener is omitted since the operating fluid temperatures do not warrant its
use. It is significant to note that
numerous fluid "users" and “suppliers" have traced the cause of particle
agglomeration caused by aging, agitation, and elevated temperatures to this
viscosity index improver. These
"spooks" or "kitties", as some have labeled them, cause
damming of close tolerance parts, small orifices, and system filters.
Careful
control of soluble water within the base stock and moisture within the system
prior to the filling operation, plus the addition of an anticorrosion additive
(calcium sulphanate), creates a compatible atmosphere for the
anticorrosion-long-term storage concept. In addition, the fluid specification requires a particle count of
less than 20 particles larger than 100 microns per 100 milliliter of fluid
sampled. It is anticipated that this
fluid will contribute substantially to the reliability growth characteristic
presented in the latter part of this dissertation.
D. SEAL,S
Available elastomeric seals were
carefully scrutinized. The one selected
for system usage (Buna-N compound) showed excellent storage stability in
approximately 150 actuators using Buna-N seals (AN6227-23, and AN6227-19) which
were manufactured during World War II and have been stored since then. Breakout forces and leakage (both external
and internal) were checked and found to be within requirements. No deterioration was evidenced. These tests are still in progress, and there
is no evidence that this rubber will not perform its intended function. Newer elastomeric compounds are also under
test.
Even with highly-reliable seals,
minimizing the number of seals (especially dynamic seals) ultimately improves
the reliability of the System. To
accomplish this, a single-ended servo actuator was selected, thus eliminating
one rod seal per actuator ( 12 rod seals per missile) when compared to the more
conventional double-ended actuating elements.
One major fringe benefit resulting from this approach was the
elimination of one servo-valve metering land (a three-way valve is used),thus
making this type valve a lower null leakage transducer and less sensitive to
the wear-leakage problem.
E. RESERVOIR
The displaceable reservoir volume
was judiciously apportioned according to the stringent weight and space
allocation. This, coupled with the
five-year instant readiness requirement, made any external fluid leakage and
even "weepage” intolerable since system precharge would be lost thus (1)
causing the high-speed pump to cavitate, and (2) allowing for the possible
introduction of atmospheric gasses into the system, producing a
"spongy" system. Both effects
are undesirable. Pump cavitation leads
to extremely short pump life or self-destruction.
Atmospheric gasses in the system
will adversely affect servodynamics as well as causing a decrease in system
stiffness. This compliance coupled with
the nature of the thrust vector control nozzle loads, (high coulomb friction to
maximum load ratio) seriously affect missile performance.
For these reasons, a bootstrap
reservoir which is spring energized for the static and initial start-up
condition is used. The bellows
reservoir body serves as the pump housing and external seals are completely
eliminated in this area through the use of a brazed bellows-to-body
design. This approach will be expanded
upon in the component description to follow.
The combination mechanical spring plus bootstrap design is used since
considerable doubt exists regarding the ability to seal a precharging gas for
five years. For the same reason, the
conventional hydraulic accumulator is absent.
F. ASSEMBLY AND CHECKOUT
The system is evacuated and
ultrasonically de-aerated fluid is used to fill the system. The small percentage of air remaining in the
system at the end of evacuation is absorbed by the fluid since the oil will
readily dissolve 10-percent standard gas volume per volume of fluid. This eliminates (1) the need for system
bleeding, (2) the possibility of contaminants entering the system during
bleeding operations, and (3) the possibility of leaving entrapped atmospheric
gasses in the system. The fluid level
is determined by observing the reservoir bellows dome deflection. Thus, the system is filled to the desired
level and bled in one operation. Figure
3 shows schematically the arrangement of equipments to accomplish the fill and
bleed procedure.
The instant readiness concept also
necessitates quick automatic checkout capability of the system. A, go/no-go pressure switch indicates the
proper functioning of the hydraulic supply.
The nozzle servo position transducer is used to monitor the actuator position
corresponding to a known position command, thus indicating the functional integrity
of the servoactuator. Reliability goals
ruled against including any but absolutely essential equipment in the system at
the silo checkpoint.
Fig 3
Evacuator and Fill Apparatus
Fig 4a Pump Flow vs Pressure (output)
Fig 4b Operational Duty Cycle Time (seconds)
A. AUXILIARY POWER SUPPLY
The Auxiliary Power Supply (APS)
converts d-c electrical energy from either a ground power source or an airborne
battery into useful hydraulic power. The
APS is essentially a constant pressure source as shown in Fig. 4 and is capable
of simultaneously supplying a specified maximum servo actuator velocity for all
nozzles.
The APS is an integrated single unit
composed of the following components:
1. Motor assembly, noise filter, and
thermistor
2. Pump
assembly
3. Check
valve
4. Filter
· 5. Remote
pressure indicator or switch
6. Fill and vent quick disconnects
7. Boot
strap reservoir
8. Pressure
transducer (telemetry purposes only)
The
hydraulic pump in the APS is directly coupled to a nominal 27-volt d-c,
compound wound motor. This motor
includes a radio noise filter which has been optimized for the specific duty
cycle. Its weight to output horsepower
ratio is about 2.3 pounds per hp.
Due to the large variation in the
system power demand, a variable delivery pump is used to minimize energy
consumption. The overall efficiency at
minimum pump flow demand is approximately 16.5 percent and at peak power output
about 63 percent.
Packaging concepts and weight
considerations induced a rapid advancement in the state-of-the-art of hydraulic
pumps. Within five months from source
selection, Autonetics received the first prototype integrated APS employing a
new Vickers high-speed, miniaturized, 905-series fixed-angle
variable-displacement pump. It
generates 3000 psi at rated discharge flow.
Inlet pressurization obtained from the bootstrap reservoir is about 85
psi. The pump is designed to operate
for rated continuous duty up to 18,000 rpm.
The overall efficiency of the 25-degree angle pump at full flow is about
82 percent at 12,600 rpm.
The remote hydraulic pressure switch
is essentially a bourdon tube-switch combination. During operation, hydraulic pressure within the coiled bourdon
tube causes it to unwind slightly. This
action opens and closes the switch contacts within certain pressure
ranges. For example, the switch is open
between 0-2800 psig and closed from 2800 to about 3350 psig, and then open
again at higher pressures.
The fill and vent quick disconnects
are of the miniature type and are only used during the initial system
evacuation, fill and bleed operation under controlled conditions as previously
described.
The reservoir is a fluid-storage
sump and during APS operation provides pump supercharge pressure through
feed-back of pump supply pressure to a differential area. The supply pressure to reservoir pressure is
inversely proportional to the respective pressure-acting areas. In the nonoperative state, the mechanical
bellows spring force overcomes the internal seal friction to compensate for
fluid volumetric changes. No external
dynamic seals are used. The stainless
steel bellows can withstand 1000 psi steady-state pressure.
A thermistor is embedded in the
field coil of the motor and is used as one leg of an electrical bridge circuit
to automatically cut-out input power in event of motor overheating during
extended ground testing of the control system.
The APS manifold provides porting of
hydraulic fluid by means of drilled internal passages. The pump output is diverted through a
modularized check valve to a common cavity which ports fluid to the hydraulic
filter and vent disconnect. The remote
pressure indicator and instrumentation transducer are located in the manifold,
downstream from the filter.
The filter element used in the APS
complies with MIL,-F-8815, Style B. When initially clean, the element pressure
drop is less than 10 psi at 3 gpm flow.
The filter is rated at 15 microns absolute.
The APS transient response
characteristics satisfy system requirements without requiring the aid of a
conventional accumulator. A time
constant of 30 milliseconds or less, corresponding to maximum acceleration of
all four actuators, is realized. The
pressure oscillations superimposed on the pump output pressure bias is within
10 percent of system pressure during the course of the transient period and are
completely damped within 0.8 seconds.
Due to its compactness, the miniature pump is easily packaged within the
reservoir, thus minimizing the inlet flow restrictions and serving as an
excellent heat s ink.
Fig 5
Auxiliary Power Supply
Figure 5 illustrates the compactness
of the APS design. The actual weight
with 10 cubic inches of oil in the reservoir is 18. 2 pounds (including a
pressure transducer for in-flight telemetry usage). Typical performance data for the hydraulic power supply unit are
listed in Table 1.
Table 1. Battery Data
|
Type: |
Silver-zinc |
|
Activation
time: |
5 sec
maximum |
|
Impedance |
Approximately
0. 02 ohms |
|
Maximum
power: |
4776 w
I& 24 v |
|
Voltage: |
27 ± 3 v |
|
Weight: |
15 lb
maximum |
|
Size: |
4 x 6 x 10
in. (maximum excluding mounting lugs) |
|
Energy: |
95 w/hr
minimum |
B. SERVOACTUATORS
Each servoactuator (Fig. 6) is rigidly
mounted to the platform structure. The
piston shaft is connected to a floating link, which in turn is pinned to the
clevis of the movable nozzle. The
primary purposes of the floating link are (1) allow for misalinements between
the platform assembly and the engine, and (2) provide for freedom-ofmotion in
the plane nozzle rotation.
Some unique features of the
servocylinder being used for MINUTEMAN are:
1. Modular
servovalve eliminates breathing face seals and bolts.
2. Torque
motor separated from the hydraulic fluid minimizes the effect of magnetic
contamination.
3. Internal
concentrically-mounted feedback transducer provides maximum mechanical and
thermal protection.
4. A-c
position feedback element gives essentially infinite resolution without wearing
parts.
5. Three-way
servovalve reduces null leakage and the number of seals.
Fig 6
Electro-hydraulic Servo Actuator
Fig 7
Servocylinder Velocity vs Force Output, for Stage II
Fig 8
Open Loop Dynamic Frequency Response
Fig
9 Combined Hydraulic System
Reliability Growth
6. Wires,
routed internally in actuator body and potted in place, provide maximum
protection against the high vibration and thermal environment.
The
servoactuators employ the single-ended (rather than the more conventional-balanced)
design, chiefly to reduce the number of seals.
System pressure constantly applied to the "small" side of the
2:1 area applies a retracting force on the ram which is opposed by the controlled
ram force from the servovalve control pressure-large area product. Thus, at a 3000-psi system potential, a net
1600-pound retracting or extending force can be realized.
The three-way actuator design is
inherently weaker in bearing load capacity than the conventional, balanced
double-rod seel-type actuator. However,
the actuator is mounted tangentially to the movable nozzle clevis when in
neutral, and with the small angular deflections required, the worst side
loading, under stall load, does not exceed 75 pounds in any missile stage.
The servoactuator force versus
velocity curve is shown in Fig. 7. The open-loop dynamic frequency response for
the servoactuator is shown in Fig. 8.
C. BATTERIES
The airborne batteries are energized
during the missile prelaunch stage, thus affording an opportunity to conduct a
last minute preflight check on the complete airborne system. In this way, reliable battery activation is
assured. This assurance carries with it
a weight premium which cannot be considered insignificant, since the system
peak to prelaunch valve leakage power ratio is 7 to 1 (Fig. 4).
The airborne batteries are the
dry-charged, primary silver-zinc type.
The electrolyte is stored in a pressurized container and, upon
activation, is discharged into the cells.
D. SEALS
Only elastomeric seals conforming to
an Autonetics rigid specification for ultra-high quality seals are used. The modularized servoactuator design employs
a minimum number of external seals: a singledynamic seal, five redundant pairs
and two single-static seals. The APS
uses a face-type shaft seal, five redundant pairs, and eight singlestatic
seals. Instrumentation seals are
excluded in the above count to illustrate the operational configuration.
E. QUALITY CONTROL
Each hydraulic component used on
MINUTEMAN is 100-percent inspected and acceptance tested. Raw material is certified and batch records
kept with each serialized part made.
Any malfunctions or discrepancies caused by material defect can be
isolated by the serialized parts identification system.
Figure 4 represents the predicted
reliability growth curve for the MINUTEMAN flight control hydraulics. Some design concepts to meet these goals
have been discussed. The nature of the
specific disciplines exercised by Autonetics on vendor items, which include the
APS and servoactuators, are outlined in Reliability and Quality Assurance Work
Statements. The highlights of these
Work Statements are summarized as follows:
1. Program
Plan. Includes complete description of
each task as well as procedures, milestones, and schedules.
2. Design
Review. Performed at scheduled
intervals. Results are to be
documented.
3. Production
Processes. All documents are defined on
internal processes and environments, specifications on parts, and materials
purchased from vendors. Document requirement
shall be reviewed, deficiencies eliminated, and a schedule for their generation
and release supplied to Autonetics.
4. Production
Control. All documents are defined for
production control specifications, control documents, and control specifications
for materials and parts to be purchased.
Document requirements shall be reviewed; deficiencies identified and
eliminated.
5. Failure
Analysis. A Failure Analysis Board of
specialists on all design and production levels shall be established. A list of known-failure modes shall be
prepared. Analysis of each discrepant
device shall be documented. A test
program to generate high-order failure modes shall be conducted. The failure modes shall be identified with
their causes.
6. Corrective
Action - Production Processes. Methods
shall be determined for eliminating or reducing defects in design, materials,
production processes, production control, and human factors. The results of this task shall be used to
improve production processes.
7. Corrective
Action - Production Controls. A method
shall be developed for eliminating or reducing process control defects The
number of discrepant items getting through inspection shall be reduced. A program shall be established for measuring
product homogeneity.
8. Evaluation
of Corrective Action. Experimental and
numerical studies to evaluate all corrective action shall be conducted and
documented.
9. Tests. Data will be obtained on specific parameters
through the generation of failure rates in accelerated environment; the results
of these tests to be used in design, for failure analysis, for monitoring,
improvement of production, and for establishment of reliability sample testing
of production procurement quantities.
10. Seller's
Program Organization. Autonetics will
monitor Seller's management of its program by establishing program management
requirement, and by:
a. An organization chart with names and
positions
b. Comments
describing responsibility and authority
c. Organization
control procedures for recording personnel and responsibility changes
d. Approval
requirements for each production step
e. Implementing
changes if required
f. Feedback
system to both Seller's and Buyer's top management of failure rate progress
11. Training. Supplier training courses shall be
established,
conducted,
and maintained to emphasize reliability-oriented programs, and incentives
covering all phases of the program.
Training should reach all levels from management to production line
personnel. Methods will be established
for motivating personnel in maintaining high-reliability standards.
12. Test
Equipment. Suppliers shall maintain a
list of failurediagnostic equipment or special test equipment which is used in
failure analysis, evaluation of corrective action, and all other tests. Also, a list of general-purpose equipment
shall be maintained.
13. Technical
Direction and Monitoring. Regular
visits will be made to suppliers by Autonetics personnel. Autonetics will station a resident engineer
at Seller's facility if required.
Supplier reliability representatives will attend the monthly technical
direction meeting.
14. Serialization. Serialization requirements shall be
established and performed.
15. Documents
and Review. Supplier will maintain in
current status all documents required by the statement of work; documentation
must be capable of being correlated to any factor relating to the production of
the hardware including design, processes, controls, tests, and failure
analysis.
These,
then, are the important steps that are expected to produce advanced equipment
which will provide highly reliable performance under extreme conditions: (1)
highly-advanced design concepts, (2) rigorous engineering disciplines, (3)
carefully stated component and materials requirements, (4) continual monitoring
of all production processes to assure attainment of even the smallest detail of
every design goal and, (5) complete documentation and regular reviews of all
design and production processes and controls, so that every possible opportunity
is realized to improve the end product.
Each of these steps is very
important. Their careful implementation
is expected to provide the highly reliable equipment required by the MINUTEMAN
design concept. Data and experience is
already being collected. Although
limited, these data indicate that the program goals can be reached.
DISCUSSION
I. PINKEL,
Chairman: Thank you
very much Lou. We now have ime for
questions.
J. KAPLAN, Arma: I was curious,
on your approach to accelerated 1 e tests, have you worked out a correlation
for example in establishing the life to accelerated G level failure and then
projecting backwards to see what the life might be under normal conditions ?
L.
PURPURA, Autonetics:
Not at this time. Incidentally, this
program is just in the midst of being funded and the modes of failure by
acceleration testing portion of the program will probably not be under way for
several months.
R. DUTZMANN,
Chrysler Missile: What swayed you to select an electrical APS as against a hot
gas APS system ?
L.
PURPURA, Autonetics:
I guess you're pushing me into some sort of a rebuttal to Mr. Deacon's very
excellent treatise on hot gas battery and inertia flywheel configurations. I can say in short, or briefly, that the
main reason for the selection, I'm sure, is basically schedule of shelf
hardware. As you probably know, the
Minuteman program has been on a rather accelerated basis and from the
standpoint of time the system that was selected being a battery APS scheme
would be almost available in months, several months, as against probably six
months to a year for any of the other types of schemes.
L. SCHWAB,
Martin: Two questions
- one, what is your reservoir spring precharge value ?
L.
PURPURA, Autonetics:
It ranges between 2 to 5 psig.
L. SCHWAB,
Martin: Do you feel
this will give you better long term
storage
capabilities than having just zero psi ?
L.
PURPURA, Autonetics:
The main reason for a slight amount of pressurization is to overcome internal
dynamic seal friction, and secondly, to have some positive internal fluid
pressure to prevent any air from getting into the system. The extent of pressurization, we wanted to
keep it quite low, in fact, just a slight positive pressure but unfortunately
Bellows spring design for our particular volume necessitated a range in this
area.
W.
MITCHELL, Boeing:
There are a couple of zigs and zags on that diagram I don't understand. One, do you have surge dampening chambers e control end of your actuators ? It looks like
there are some little chambers stuck down below.
L. PURPURA, Autonetics: This is the servo valve. This is drawn incorrectly. These lines should go accordingly. This, essentially, is to indicate the
position transducer. It is mounted
concentrically to the piston and this is just the aft section of the variable
differential transformer, the fixed portion.
The core, or the movable portion, is fixed to the ram itself.
W.
MITCHELL, Boeing: The
other question was, you apparently have two different actuator designs.
L. PURPURA, Autonetics: No. Two different
people made this drawing.
C. CANNON,
Lockheed: I just
can't resist making remarks about reliability again because we had the original
comments about the best way to get reliability is in the design of the system
and yet we sit here and look at a simple schematic diagram of what should be a
fairly simple hydraulic system and really not much advance in the state of the
art over what we had 15 years ago, and yet to get reliability we're designing
assemblies and plants that are better than the best hospitals in the country
and we're setting up the companies, reorganizing them, telling them how to
manage and run their company. It seems
like weire doing everything to change the world around designing reliability
into our system, and I think what we've trapped ourselves with is the electro
hydraulic valve. I think that's the
only thing that's really different in the last 7 to 10 years. It's a thing that just really can't stand
contamination and of course if it fails it's always catastrophic. It fails hard over or gives us a serious
signal but basically when we look at our actuators and our pumps and our other
equipment, we've had all these other gadgets for years and we've had small
orifices in systems, and we've eliminated fittings, which is good, by
manifolding but I think probably we need a breakthrough in an electro hydraulic
valve field. Maybe if we hadn't come
around with these nozzles and flappers we would have used an old spool valve
with a motor or something on the outside driving it like a manual servo valve
and maybe we would have saved a lot of this cleanliness and big worry about
keeping the system reliable from that standpoint.
L. PURPURA,
Autonetics: I concur Mr. Cannon.
Unfortunately, as I stated previously, the program is one of
acceleration from the standpoint of what do we know about existing hardware and
what has to be done to more or less extend the reliability of existing
hardware. I think the industry is aware
of the existing problems and how to approach these problems and what sort of
controls are needed to combat the contamination sensitivity of valves. We're prese using two types of valves, a jet
pipe type and also the flapper nozzle valve.
To date, we have not had
contamination problems for several reasons.
We insure adequate fluid cleanliness before installing the fluid or
injecting fluid into the system. We
have rather rigid cleaning controls of the platform which is a very big
bug-a-boo obviously with all its long drilled passages as well as the
integrated hardware. But I would say
that the greatest weaknesses of hydraulics is probably external leakage. We have a very limited system volume and from
the standpoint of intermittent checkout testing in silos and shear problems of
shaft seal leakage and this sort of thing.
I'd say that this is our most critical area from a design point of
view. I believe Mr. Hecht bore this out
yesterday. Everybody is concerned about
valve contamination problems but the problem has not been very severe. I think the main reason it hasn't been
severe is that everybody is aware of the sensitivity a valve is to contaminates
that we all are cleaning up our house and insuring our systems do not include
either contaminates built into the system or added to the system. I might add that one of the concepts of the
Minuteman is no field maintenance. I
mentioned the use of these disconnects.
These particular disconnects are only used at Autonetics under clean
control room environments and for filling and bleeding and this cap is just a
redundant cap on the check valve, or disconnect, and if any failure should ever
occur in the field, the whole unit, the whole platform assembly is removed from
the stage and sent back to Autonetics and disassembled. As I indicated in the paper, in recent Air
Force studies, the majority of missile aborts were accomplished by human
factors or human tinkering with the particular system and this is what weire
trying to design around. I hope I
answered your question Charlie. I know
some of the problems.
C. CANNON,
Lockheed: I really wasn't putting the comment to you. I just put it in because it keeps hitting my
mind over and over that we haven't really designed reliability into the
systems. We've redesigned all the
factories in the country and all the procedures and everything else to get
contamination out of a system rather than doing something about the system to
work on contamination. I think maybe we
all ought to concentrate our efforts on that.
I know I'm guilty of it too.
I've done the same thing. I've
made them change around all the equipment in the factory and procedures and
made everybody else jump. I think our
big breakthrough has been in manifolding components and getting rid of joints
and fittings.
I. PINKEL,
Chairman: We're pressed for time this. morning. I appreciate that many of you would like to continue this
discussion. We'll allow one more
question.
M. OGMAN,
Convair: 1 believe you mentioned that you were not using an accumulator
in the system. What has been your
experience with pump pre s sure pulsations ?
L. PURPURA,
Autonetics: As 1 stated in the paper the pressure rippled to a transient
condition. It's approximately ± 10@o
from a system bias pressure say of 3000 psi.
The time constant for this particular unit for two-thirds of peak flow
is approximately 25 to 30 milliseconds.
I. PINKEL,
Chairman: Our next speaker is Dr. W. W. Chao who is Director of Research and Development for Vickers
Incorporated. Dr. Chao received his
doctor's degree in mechanical engineering from M. I. T. and is a member of IAS,
ARS, ASME, SAE, Combustion Institute and Sigma Psi. His past associations have included Curtiss Wright and Bell
Aircraft. At Bell Aircraft he served as
Corporate Representative on SESA, and I hope he tells us what this is, and the
Liquid Rocket Committee of AIA which is the Aircraft Industries Association, if
you don't know what this one is. They
call themselves now the Aero Space Industries Association. Also, at Bell Aircraft he was responsible
for the research design and development of propulsion systems including the
second stage liquid rocket engines of the Discoverer Satelite, the high energy
system for Mercury and the X-15 airplane.
The title of his paper this morning is, "Attitude Controls Development
at Vickers for Space Vehicles and Missiles." Dr. Chao.
****O****
Looking Back
It turned out that the Stage I
versions shown in the prior photos were never flown, they had to be resized --
stage I was increased to be 50% more powerful.
Each stage had four tilting nozzles
and the peak demand had been calculated on maximum pure Pitch or pure Yaw. (Roll was achieved as Pitch + or -
Yaw). When Clarence Ashe, Ron Frazini
and George Leonard joined me we decided to start from scratch; we read all
correspondence and did our own sizing calculations. We found that the worst case is when Pitch and Yaw must move
simultaneously in 45 degrees to pure Pitch or Yaw. We also found that servo valve leakage during standby was not
taken into account, this leakage is at very low efficiency and thus a
significant battery drain. After
carefully checking our numbers I wrote a note to Ray Curci, head Project
Engineer, who was at the time tied up in a meeting: “There is a strong possibility
that the first flight will fail. The
system is undersized.” His secretary
took the note to him and a bit later he was at my desk asking “what the hell is
this about!” Ray a good engineer was
soon going over our calculations. He
said, “I’ll try to stop first launch, you be ready to move out fast on what we
do need. The stage I system was changed
to 7.5 HP, and the actuator force outputs increased by 25% and battery capacity
increased 50%. It was a remarkably fast
turn about – something we could not have done later with added red tape.
The resultant APU’s looked much the
same as fig 5, except the motor envelope and radio noise filter relative size
was less.
The new actuators looked
different. The lump at the back of the
actuator in Fig 2 & 6 is the servo valve, this was moved up closer to the
mount. I said I didn’t like that valve
at the end of a cantilever, that I thought it would be better moved up to the
other end near the mount. I was asked,
“do you know it will be better off there?” “I said no, but if I wanted to cause
a valve to fail under vibration I put it at the end, as it now is, there would
be some frequency at which it would see more vibration amplitude – and we don’t
have time to prove it by test.” After
others thought it over, they agreed, and it was moved to the front end of the
servo.
Fig 3 shows how we solved the
dilemma on bleed and fill the system.
The “systems” people were supposed to find a way to bleed and fill the
system. Two of their young engineers
came to me to see if I had any ideas. I
was busy at the moment and said, let me sleep on it. The next morning I had a method worked out and arranged for our
outside lab vacuum pump to be brought into inside lab along with some other
items. I made a sketch of how to
evacuate the new oil container and the NCU cavity, then when all entrapped air
was removed to release it into the system.
Bleeding had always been done by flowing oil through passages to carry
out the air. I knew the vacuum pump
would suck the air out but I also knew from test done by Bob McCoy using a
plexiglass actuator that oil contains a lot of entrapped air which will come
out of solution if pressure is reduced.
The fellows later added the use of a sonic agitator to help free the
bubbles. So far as I know the method is
still used at the AF repair depots.
The actuator design was unique. If you look close at Fig 2 you will see that
the body of the actuator bolts to the frame.
There is and “extensible link” on the end of the actuator shaft. When the actuator moves in or out and the
rocket nozzle tilts, the link moves up or down to correct for linear vs curving
motion.
This permitted using an actuator
with “one shaft” coming out the front and none out the rear. This permitted putting the position
transducer inside the actuator, fastened to the back of the cylinder and
encompassed by the hollow piston shaft.
This was not only protected and reliable but classic binding of position
sensing devised was avoided.
Wires extending from the
servoactuator in fig 6 were threaded through drilled passages in the frame
which served as electrical conduits, there were no electrical connectors –
wires were soldered to electronic boards.
Frank Lettang was given the task of
how to protect the Nozzle Control Units from the 2000 plus degree gas coming out
of the tilting nozzles. Frank came up
with a light weight rubber material that could be “painted” on a completed
assembly. Thus the systems ran cool
though next to the fire.
Minuteman I was an overwhelming
success. But technologies kept changing
choices available to engineers.
Minuteman
II would change to Secondary injection
for Stage II as the means of attitude control.
Also Flight Control and Guidance System Electronics would change to
better semi conductors. Minuteman I
used discrete parts, Minuteman II would use small scale integrated circuits
made to “Minuteman parts standards”.
Minuteman
III would add multiple warheads, a Post Boost Propulsion System, secondary
injection for stage III and medium scale integrated circuits for electronics.
Missile
X would shift to gimbaled booster nozzles and hydraulic servos. The tilting nozzle system consumed too much
envelope, reducing potentially available thrust. Secondary injection weighed too much as compared to the new
gimbaled nozzles.
Though
very successful, they system described here was removed from the inventory.
Today
the once very real threat from Russia has drifted into a lesser but uncertain
threat.
Many
nations now know how to make ICBM capable missiles – if a leader wishes to
squander his nations assets.